الفهرس | يوجد فقط 14 صفحة متاحة للعرض العام |
المستخلص Abstract: The current study aimed to investigate parameters estimating methods in light of the Item Response Theory, specifically the unidimensional logarithmic model and the graded response model. Two types of items, dichotomous items and polytomous items, were generated using the WinGen-3 program. Twenty dichotomous items and fifteen polytomous items were generated for sample sizes of 500, 1000, and 1500. The item parameters resulting from the simulation were then estimated using the unidimensional logarithmic model and the graded response model, using the following methods: Maximum Likelihood Estimation (MLE), Maximum A Posteriori (MAP), and Expected A Posteriori (EAP). The study concluded there is no significant difference in the precision of estimating parameters for both dichotomous items and polytomous items across estimation methods at sample sizes of 500, 1000, and 1500. Additionally, the precision of individual ability estimations for dichotomous items at sample sizes of 500 and 1500 showed no statistically significant difference. However, there was a statistically significant difference in the precision of individual ability estimations for dichotomous items at a sample size of 1000, favoring the MAP and EAP methods. Moreover, the precision of individual ability estimations for polytomous items varied significantly across different sample sizes, favoring the MAP and EAP methods. |